The identification of freezing-thawing soil hydraulic properties is the basis and key problem of studying soil water movement. Current identifying methods of the properties have many limitations, such as big error, high uncertainty and low operability. Thus, the identifying method of freezing-thawing soil hydraulic properties was researched in the research. The estimating parameters of soil freezing characteristic curve were calculated by classical statistical method. Targeting at the maximum absolute difference between measured and simulated value of unsaturated soil hydraulic properties, the particle swarm optimization algorithm was used to estimate the optimal values of soil water characteristic curve, unsaturated hydraulic conductivity and specific water capacity. The numerical simulation of freezing-thawing soil water movement proved that the measured and simulated value of soil unfrozen water content had high degree of agreement. The identifying method proposed in the research is effective and feasible.